央行微观调查数据适合作为不良贷款率的预测指标吗——基于MIDAS模型的研究

发布时间：2018-03-24 17:28

本文选题：央行微观调查数据　切入点：不良贷款率　出处：《当代经济科学》2017年03期

【摘要】：相比于传统宏观经济指标,央行的微观调查数据是否更加适合作为我国商业银行不良贷款率的预测指标?本文首先提出了利用微观调查数据进行预测的相关理论,接着基于适合解决混频数据预测问题的MIDAS模型进行了实证研究。研究发现:央行微观指标在平均样本外预测误差等多个方面均优于传统的宏观指标,同时人民币名义有效汇率指数、基金及理财投资意愿比例、房价过高难以接受比例等指标尤其适合作为预测指标。本文的结论是:央行的微观调查数据更加适合作为不良贷款率的预测指标。
[Abstract]:Compared with the traditional macroeconomic indicators, whether the central bank's microcosmic survey data is more suitable as a commercial bank non-performing loan ratio prediction indicators? In this paper, we first put forward the theory of using microcosmic survey data to predict. Then based on the MIDAS model which is suitable to solve the problem of prediction of mixing frequency, it is found that the micro index of central bank is superior to the traditional macro index in many aspects, such as the average prediction error outside the sample, and so on. At the same time, the nominal effective exchange rate index of RMB, the proportion of funds and willingness to invest in financial management, The conclusion of this paper is that the microcosmic survey data of the central bank is more suitable as a predictor of non-performing loan ratio.
【作者单位】：南京大学商学院;中国人民银行连云港中心支行;
【分类号】：F832.4

【相似文献】